‘Oh, Jo! You can’t see that
real life is not like riding a horse!’:
Clients’ constructions of power and metaphor in therapy
the Research. 
This paper is based on my Ph.D thesis looking into
'clients' and 'therapists' constructions of
metaphor and metaphor use in therapy. The research itself involved a
version of discourse
analysis of accounts that resulted from a number of semi-structured
interviews with participants,
as well as from the transcript of a focus group. I recruited a total of
16 participants (12 women
and 4 men), 8 of those were therapists, including a Dramatherapist and
5 were clients to
therapy. An additional three participants were recruited for occupying
the double positioning of 'therapist in training' and 'client' to
(i.e. hence having access to discourses from both positionings) and
took part in a focus group with the aim of exploring meaning
action. All names that are included in the excerpts in this paper are
pseudonyms, in no way
indicative of ethnic origin. The decision to recruit a Dramatherapist
reflects my personal
interests and familiarity with the field and was aimed at exploring
dominant discourses in this
area, while in no way implying the account is 'representative' of
were recruited solely on the basis of their availability. Categories
were in all avoided, unless as
made relevant by the participants themselves in talk (e.g. model of
practice, ethnicity). I did
however make an effort to represent a wide range of therapeutic
discourses, through recruiting therapists from as wide a range of
models of practice as possible. Models represented were
various constellations grouped under an Integrative or Eclectic
approach (e.g. combinations of
existential, Cognitive Behavioural Therapy, cognitive, Cognitive
person-centred, psychodynamic, Jungian, systemic, humanistic, core
theory), as well as the
psychoanalytic approach. A number of therapists were recruited from two
NHS Trusts. Due to
the nature of the topic, an attempt was made to recruit experienced
therapists, where possible, to
counter-balance issues such as the novelty of the practitioner or a
lack of familiarity with
metaphor in therapy and its uses, both of which may impact on the
richness of data. I therefore
approached a number of experienced former colleagues, who I knew had an
metaphor. This was an attempt to ensure that data would be sufficiently
rich by only including practitioners with an expressed interest in
metaphor work in therapy. I used a 'snowballing'
technique, where the research was advertised by word of mouth. Clients
were recruited in four
ways: 1) via posters placed in the psychology department at City
University aimed at students
training in Counselling Psychology on the MSc. and Post-MSc. courses
(i.e. for the dual
positioning of client/therapist focus group). 2) I also placed
advertisements in alternative health
centres in my local area, which offered counselling and/or
psychotherapy. 3) I circulated the advertisement via the email list of
the Critical Mental Health Forum group, notifying participants to the
Forum of my research.
Utilising a critical realist and constructionist
perspective, where no claims as to the ontological
status of metaphor were made, I was instead interested in the
participants' discursive resources
in accounting for the phenomenon of 'metaphor' and 'metaphor in
In order to achieve
this I conducted a number of semi-structured interviews with therapists
and clients to therapy,
interviewed separately, as it had not been possible, for ethical
approval reasons, to work with
interactional data (i.e to conduct interviews where therapists and
clients discussed metaphor
together). The purpose of these interviews was to collect discursive
accounts that pertained to multiple constructions of metaphor and other
discursive considerations, such as the discursive
and rhetorical strategies participants enlisted. In addition, I wanted
to explore the subjectivities
of the clients in relation to metaphor use in their therapy. In order
to make up in part for the lack
of interactional data and to observe co-construction of meanings in
action, I ran a focus group in
which participants were invited to contribute their views from a double
position of being both 'therapists in training' and having been
to therapy at some point in their lives. Fourteen texts were
subsequently transcribed in a style that emphasises readability rather
Analysis of texts
The discourse analytic method represented an analytic decision based on differing research interests in terms of each research group (therapists and clients) and it was finally crystallized in the research as representing a difference in emphasis between the two methods employed. In discursive psychology, the focus of the analysis is on the fine grain rhetorical functions of the discourse-how people use language and what is accomplished by the employment of certain discursive strategies (e.g. disclaiming, footing, extreme case formulation etc.). Attention is therefore paid to the action orientation of talk, which represents multiple ways of reading talk (e.g. Edwards, 1997; Edwards and Potter, 1992). The context is always of paramount importance (i.e. readings are local). For example, variability in one account in terms of interpretative repertoires, suggests that objects are constructed through language and that discourses are inherently dilemmatic (Willig, 2000). In the therapist's accounts, the focus was therefore on the functions performed by language, under the assumption that there is 'stake' in the way the therapists are positioned by the interviewer in relation to the topic of metaphor use in therapy (an 'expert' positioning). As a result the action orientation of the texts was explored, as well as the rhetorical effects of language. The only exception to this was the Dramatherapist's interview, which, being more of a narrative (monologue) was also analysed with a Foucauldian emphasis. Foucauldian discourse analysis is more reconcilable with an applied framework. For the client's accounts, the Foucauldian version permitted an exploration of the extent to which expert therapist discourses were affirmed, resisted or altered by clients. It also enabled the exploration of multiple and varied constructions of metaphor' and constructions of subject positions. In addition, some of the resulting political, theoretical and applied implications were considered. Finally, the focus group's combined version looked into all of these analytic aspects.
'Power' in discursive psychology is dually
theorised as the pre-discursively defined inequality
of relations such as in the case of institutions (i.e. institutional
power) and as an interactional
quality through the strategic employment of discursive strategies
(Jaworski and Coupland,
1999). The concept of power is also particularly compatible with a
Foucauldian approach to
discourse analysis, which suggests that institutions, like therapy and
mental health organisations
are structured around and reproduce power relations (Parker, 1992). As
Harper puts it: "it's
about who gets to speak and who gets to choose the words we use to talk
(2002a, p. 9). Another view that is
currently challenged is that any
therapy talk is helpful. Talk
can just as easily be oppressive rather than liberating or helpful if
it is the wrong kind of talk, in other words if it is pathologising or
blaming (Harper, 2002b). Metaphor as an
feature of talk potentially has the same implications. Power inequality
as an inherent aspect of
the therapeutic relationship is now widely accepted. As such, it cannot
be denounced, but must
be linked to responsibility. In White
and Epston's (1990) words:
The empowerment of the client, as the realisation of the
capacity for oppressed groups to have
access to power (Bhavnani, 1990), is
therefore an important task for
therapy. Anderson and
Goolishian (1992), for example have tried to redress the balance by
advocating a way of
working where the client is seen as the expert in their own experience
(also Anderson, 1997)
and Narrative Therapy also denounces the 'knowing expertise' of the
therapist (e.g. White and
Epston, 1990; Cheshire et
al., 2004). As certain discursive and
signifying practices perpetuate
positionings that compromise the client's egalitarian position, it is
only by becoming aware of
power issues in therapeutic practice that we can strive towards a
politically aware  and ethical practice. It is also
as some client participants to this research suggested,
therapists can be themselves oppressed by institutions and practices
over which they have little
control. This point is made in the following client quote:
A Discursive Psychology approach to the material would
have also been possible. From this
perspective, clients' constructions of power could have been
understood, for example, as
disclaiming devices for personal agency and responsibility. Instead, my
interest lay in exploring
the extent to which clients' discourses affirmed those of the
therapists or whether they diverged
from them. For this reason, I want to emphasise here that the status of
the texts has a realist
ontology, as client accounts are taken at face value, as authored
accounts of self-expression
(Willig, 2004), in order to explore
issues of power in therapy.
Constructing metaphor and power in therapy
In the examples that follow, clients construct metaphor in ways that have political implications and explain how certain uses of metaphor by the therapist work as metaphors for power positions. For example a particular physical set up of the therapy space, certain scripted behaviours and institutional practices only enable certain positionings for client and therapist and restrict access to others.
Power-laden uses of metaphor
a) Using metaphor as a silencing device
Clients described the therapist using a client- or therapist-generated metaphor in a way that acted as a silencing device for the client. In these cases, the client's worldview was closed down and denied in favour of the therapist's (e.g. 'can't you see that real life is not like riding a horse?'). In this way, the therapeutic relationship is itself constructed as a power struggle, like a game of 'tug of war'. In the first excerpt, the client, Jo, uses 'riding a horse' as a metaphor for life. She describes her therapist's response as un-helpfully obliterating that metaphor and worldview in favour of an alternate perspective, that of the therapist's:
b) The high-jacked metaphor
Client-generated metaphors were at times constructed as 'high-jacked' by the therapist, who put a different meaning on them. Below, Linda draws on a discourse of ownership ('that's mine, give it back'). Her positioning as a helpless child also evokes a family discourse. She uses action verbs ('taking your metaphor and running away with it') to describe how metaphor is snatched way from her like an object, while she positions herself as passive in the interaction. Her only reaction is to shout back ('Come back! That's mine!'), as she represents herself as incapable of reclaiming the object (i.e. metaphor). Instead, she has to plea for its return. The extract evokes playground bullying. She portrays this use of metaphor by the therapist as power-driven and unhelpful:
The example below positions the therapist as powerful,
in that he or she has the right to ask
probing questions. The legitimisation of probing questions by the
therapist, in turn positions the
client in a pathologising way, a position she then has to work hard to
defend against and
c) The signifying environment
The arrangement of the therapy environment and certain institutional practices like note-taking, act as meta-communicators and specify particular power relations between therapist and client, which govern and regulate the nature of the interaction by making only certain positionings possible. Furthermore, these discursive practices produce and re-produce these rules. In the first example below, a client describes, how the set up of the therapy room already positions the therapist as 'expert' and 'powerful' and the client as 'passive' or 'emotional' by describing the room furniture not as arbitrary arrangements, but as powerful signifiers of specific power positionings. For example, by having a bookcase behind the therapist (knowledge and power) and a box of tissues and a bin (a receptacle perhaps for emotional rubbish) next to the client. The client describes this as a dictation of the role she is expected to take ('I'm supposed to sit there and cry'):
In the second example, again the physical space and the
objects contained within it spell out
the differences between therapist and client: the therapist is
obviously affluent, the client by
implication is forced into a comparison through which she emerges as
disadvantaged. The use
of 'ostensibly' is pivotal here in establishing a conscious and perhaps
intentional display of
power positionings for therapist and client. Other elements also act as
signifiers such as the
The client from the first example, however, constructs
herself as 'reactive' to this physical
arrangement of the space and the power positions it spells out, by
refusing to cry and by having
an urge to upset this status quo:
She further constructs herself as 'assertive', as an
educated person. She is able to stand up for
herself, question the positionings and the input of the therapist. She
describes this position as a 'challenge' or 'threat' to the therapist.
is possible, therefore, to challenge power positionings
and to reclaim some of that power, but it takes an assertive, secure
and educated person to do
Another form of exerting power in therapy is on an institutional level. Clients referred to the following means of power-exertion, which have become routine means of control of the client and the process of therapy (i.e. they are normalised in therapy contexts as part of the process of therapy in organisations, a kind of 'natural order of things' argument):
i) scripted behaviour
ii) note-taking and
'Withholding' refers to those styles of therapy, usually described in psychodynamic approaches that advocate the therapist's role as that of a 'blank slate' for the client to project onto. These approaches also tend to favour a low level of personal disclosure on the part of the therapist.
a) Scripted behaviour
Clients constructed certain rules that seem to govern interaction in a therapy context as oppressive. This includes the therapist 'going to fetch' the client (a common practice in counselling within institutions) and leading the way to the counselling room. This scripted behaviour was described as 'unnatural' and 'restrictive'. The power is also firmly placed in the hands of the therapist as the initiator of the therapy interaction:
Perhaps it is a useful point for reflection, the
implication that what counselling psychology
constructs and normalises as giving 'permission to cry' or not
establishing conversation with the
client outside the therapy room as 'demarcation of the therapy space',
are instead constructed by
clients as oppressive practices.
Note-taking, though not uncommon in therapy, is
constructed here by clients as another form
of oppressive practice. Notes represent a fixed sort of narrative, that
is 'there' and which can
have damaging implications (i.e. different sorts of people might see
it, it cannot be altered, the
client has limited or no access to it and doesn't have a say in it,
etc.). In the first excerpt, Jay
makes a dichotomy of power, between doctors (referring to a
psychiatrist) and clients. The
former are 'all-powerful', the latter 'nobodies'. This is a clear
of what Potter (1996)
calls 'category entitlements', by which speakers invoke membership to a
when building up an account as sufficient to warrant category-specific
they are not 'natural' but need to be 'achieved' or 'worked up'). In
case to be a doctor equals
power, to be a client equals being powerless. Access to notes is only
restricted to the doctors
and the therapeutic relationship is therefore constructed as un-equal:
Below, Rhona also refers to note-taking, therapy in the
therapist's own house, as well as
therapeutic techniques as patronising, since the client is positioned
on unequal footing and can
also perceive therapeutic techniques, described here as a sort of
('she was using a technique to see what I would do'):
Clients also constructed a helpful therapist to be 'directive', as opposed to 'withholding'. The latter was referred to as an unwanted and alienating feature of therapy. This referred specifically to therapists who 'sat back' and it was "like to just talk to a blank" (Rhona, 535). By contrast, metaphor is described as a realm where a 'real meeting' can take place:
In the next example, the 'withholding therapist' is
constructed as antagonistic, while the client
becomes 'reactive' through the use of silence (see section on clients
'acts of resistance') and the
relationship is described as reaching an impasse. The content, however,
is contradictory. While
the client seems to be saying she does not want guidelines or answers,
she does want something
In the next excerpt, Elaine, in talking about her
experience with her two Dramatherapists,
raises an interesting point about the counter-intuitive ways in which
'giving direction' ("they
gave the instructions, now you do this scene, now you do that scene",
Elaine, 405) is equated
with equality (c.f. doing nothing with power):
Where the therapist was described as 'withholding', the client's metaphor was also described as 'wasted'. Here, Jo describes how the metaphor she used in one session was ignored, only to be referred to again by the therapist in the next session, by which time Jo felt the metaphor had been 'wasted':
The 'unspoken metaphor'
The 'client' was occasionally described as 'circumspect' in relation to the use of metaphor. This construction is particularly evident in Jo's interview. Particularly, where the therapist's knowledge of the client's history of 'psychiatric illness' is concerned:
Where metaphors were judged by the client to be
'out of the norm', 'disturbing' or 'bizarre',
they were withheld and not shared with the therapist. Similarly to the
withholding therapist, the
client also becomes withholding. The judgement of metaphor was done by
reference to 'normality' and to everyday 'conversations':
Jo constructs a 'timeliness' of metaphor as important for metaphor work. Indeed, she suggests that a metaphor that is not taken up in the moment is a wasted metaphor. She further suggests that for the therapist to merely repeat the client's metaphor without exploring possible meanings is also a waste ('I had known that anyway').
Acts of resistance
Clients positioned themselves in talk as both 'helpless'
and 'assertive'. For example, they
represented themselves as 'helpless' when they were fighting over
ownership of metaphor (as in
the section 'the hijacked metaphor') and associated meanings, but also
as 'assertive' by
attempting to resist unequal positionings in the therapeutic encounter.
Simpson (1977) states
that while individuals can be forced to conform to specific forms of
behaviour, there are still
"possibilities for resistance" (p. 221) The latter positioning is
demonstrated in the interaction
between two members of the focus group (therapists in training but also
clients to therapy) that
follows, who speak from a client's position:
The capacity for assertiveness however, was placed in a context of age, as well as professional status (i.e. being a psychologist)(c.f. power/knowledge). So, education and age were constructed as assisting an 'assertive' subject position, in order to voice dissent and make corrections in terms of the meanings of the client-generated metaphor as misrepresented by the therapist. In the absence of these conditions assertiveness was deemed difficult:
Though none of the client participants talked about physically leaving the therapy space when positioned in unfavourable ways, they nonetheless talked about the possible employment of other discursive and interactional resistance and protest strategies to these positionings. In this section I explore a number of 'acts of resistance'.
a) Making personal comments and asking
Mastache (2003) reports acts of successful client resistance to power positionings, such as asking the therapist personal questions (op. cit). In this research, the strategy of making personal comments, we are told in the extract below, is unyielding of positive results, such as the egality of the therapeutic relationship, as the client attempts to position both herself and her therapist as equals. Though this construction can also be viewed as a device that manages stake on the part of the client who escapes issues of responsibility by reversing the tables on the therapist, we are told here that as a strategy to bring about egalitarian equilibrium it is unsuccessful, as the therapist is disproving of it. In the extract that follows Jo makes a comment on the therapist's tan and ponders a question that concerns the therapist's holiday, just as reported by Mastache (2003). The difference is that while in Mastache's research the therapist reinforces a democratic climate and answers the client's questions, even interweaving personal narratives (op. cit), the therapist in my research is said to be 'disapproving'. Jo evokes a family discourse where she positions herself as a child and the therapist as a parent (e.g. "you're not allowed", "she was very disapproving") who needs to ask for permission for her actions, rather than, as an adult, initiate a discussion:
Another discursive strategy of resistance is the persistent recapitulation of issues in therapy as a response to the therapist's withholding. In the next example, Jo says she persisted in her attempts to elicit a response or feedback from her therapist:
Repetition as a strategy of resistance is also evident in the next quote:
c) Giving feedback
A strategy of giving feedback to the therapist is an attempt to redress power imbalances. In the extract below, the therapist is said to resist a new order of things when the client offers unfavourable feedback about the therapeutic relationship by ending the conversation ('oh thanks for telling me that'). Her refusal to engage with this issue and withholding stance can be interpreted as acts of resistance to the client's resistance:
and then at the end I told her this I
told her I
really communicated, we haven't really communicated
and she just said oh thanks for telling me that and that
was it really (Jo, 386-391).
Another act of withholding, this time on the part of the client, also suggests that by rejecting everything, Elaine resists being positioned as a responsible and autonomous agent. She does not want answers, she does not want guidelines, she does not want help and she does not want to be the one to ask for these things either:
d) Going defensive
Another withholding strategy is for the client to go defensive. We are not told exactly what this entails, however it is resisting being positioned in an objectifying way rather than in a relational way, such as in the case of a passive therapist who, acting like a mirror, necessitates a very active client:
I suppose having a bit of a therapy
myself I was
rather aware of the techniques and it made me feel a
little bit unsafe the way they make themselves very bland
and try to get you to project and I think I went very defensive
An egalitarian strategy of resistance to an otherwise unequal balance of power, where the therapist pursues a line of a misinterpretation of metaphor, is for the client to 'speak up'. As in 'giving feedback', the 'assertive' client is a client who does 'correcting' as a strategy of challenging and redressing power positionings, including the misconstruing of meaning. Through this strategy, the client exercises control over meaning:
Implications for Therapeutic Practice
Reflecting upon these power-related aspects of metaphor, I want to advance an argument of 'usefulness', which consists of a utilitarian angle, offering practical recommendations to therapists. The discussion that follows is concerned with this utilitarian argument and also advances a political agenda for therapy, which has as a main therapeutic aim the empowerment of the client. By 'empowerment' I mean both within the context of therapy (i.e. assuming therapy relies inherently on an imbalance of power) and beyond, in a social context. The issue of power in therapy has been well documented in the therapy literature (e.g. Killikelly, 1995; Webster, 1995) and the issue of power imbalance as endemic to the therapeutic encounter is now generally accepted. Therefore, Carswell and Macgraw's (2001) suggestion of a political agenda for therapy is timely. The agenda can choose to be politically empowering and subverting by effecting change or politically conservative whereby it enforces and supports societal structures. The paper attempts to feed into this kind of political and empowering agenda. Carswell and Macgraw (2001) also comment on the double bind therapists may find themselves in, when even as they work to empower clients they are themselves dis-empowered by the institutions in which they work. Aside from issues related to the organisational and institutional environments as contexts of therapeutic work, often conservative and resistant to a political agenda for therapy, the authors also provide the example of therapeutic stories, usually narrated in the therapy literature from the therapist's point of view, while the client as a source of information for the therapeutic relationship is ignored. Social constructionist thinking and approaches to therapy identify familiarity, democracy and equality as desirable characteristics of a therapeutic environment (Mastache, 2003; Anderson, 1997; Hoffman, 1990). Mastache (2003) has looked at the details of therapy talk and explored ways in which the above democratic features become visible in 'institutional talk' when this overlaps with 'ordinary talk' (i.e. such as 'holiday talk' initiated by the client). The overlap between the two ways of talk is concerned with a change in the relationship of power between therapist and client (Mastache, 2003). This represents a departure from the view that the two kinds of talk are different and almost mutually exclusive (Silverman, 1997; Drew and Heritage, 1992). Discursive displays such as the therapist's 'self-disclosure', the client's' self-diagnosis' and the client 'asking questions' are all examples of unexpected ways of talk in therapy, which promote a more democratic and egalitarian environment (Mastache, 2003). These feature increase with a certain kind of therapeutic environment. These findings that suggest the client is active also oppose the traditional view that clients are passive recipients in therapy or have a position limited to telling the therapist their problems (Mastache, 2003). Traditionally, the questions of the client questions can be construed by the therapist as 'resistance'. By answering questions directly, providing some detail, the therapist is promoting an egalitarian environment for therapy.
Clients to therapy have reported that familiarity,
feeling comfortable and relaxed and a relaxed,
informal environment are all important aspects of therapy (Mastache,
2000; Anderson, 1997).
The present research supports and adds to these findings. In this
research clients represented a
therapeutic environment that creates, promotes and sustains power
relations as unhelpful and
undesirable (e.g. the signifying environment). Furthermore, metaphor
use in a therapeutic
context has been largely represented as an arena for client-therapist
conflict and inequality.
Where meaning was arrived at non-collaboratively the therapist shuts
down its exploration
rather than opening it up in an act of co-construction, uses metaphor
as a silencing device or imposes his or her own interpretation of
meaning on the client's metaphor, thereby making it'
something different and unrecognisable for the client. This, we are
told, is the process that takes
place instead of processes of clarification or co-construction of
meaning (e.g. see the
high-jacked metaphor). The
physical set-up of the therapy environment
is also important. Rather
than promoting power inequalities (e.g. such as in the signifying
environment), due attention
and thought might instead promote the relaxed, informal and egalitarian
atmosphere clients say
Not all metaphor use has been discussed within a context
of un-helpfulness or oppression.
Other uses of metaphor as an educational device (e.g. explanatory of
self-help in the form of books, etc.) or where therapist and client
were equal, unknowing parties
in the co-construction and open exploration of meaning, have been
constructed as helpful and
empowering. As metaphors make certain worldviews available and have
positionings, the therapist could be aware of those and use metaphors
(i.e. in the case of
therapist-generated metaphors) that do not compromise the client by
offering unfavourable or
unequal subject positions (e.g.
the re-parenting metaphor). Moreover,
open-shaped metaphors that reflect a sort of progress or movement (e.g.
journey, metaphors of growth or expansion,
etc.) or spatial metaphors (e.g.
metaphors of landscapes) offer many
possibilities and could be
preferred ways of using metaphor to describe for example the process of
therapy or a
psychological state. I am attempting here to draw attention to the
importance of the therapist's
awareness of the implications
of metaphors (i.e. the worldview and
positions offered within
that, as well as their desirability or appropriateness), rather than be
prescriptive as to how
metaphors could be used or
indeed should be used.
The training of counselling or clinical psychologists
could also reflect the ever-presence of
metaphor. A number of therapists in the research commented that they
had mostly become
aware of metaphor use in therapy in their own practice rather than in
their training. Where
metaphor work in the training was mentioned, it was in the context of
having a small, peripheral
role if any, or else being discussed in relation to the training itself
(e.g. metaphors of group
process in training, issues of transference/counter-transference of
students and lecturers, etc.),
rather than ways of working with metaphor with clients in therapy.
Moreover, if all language is metaphorical, but metaphors become reified
as versions of reality as we speak them (i.e.
through the reification offered by language itself), as therapists we
would surely benefit from an
awareness and a (joint) decision as to what versions of reality is
helpful or appropriate to
advance with our clients. The client's own direction in this would be
valuable. In the act of
co-construction, different metaphors can be created and explored for
the fit they provide with
the client's worldview, as well as others be created, which might offer
other, alternative ways of
looking at the world, which the client may find helpful or useful.
Where metaphors don't fit' or
are outgrown, they are simply and lightly discarded in favour of
others. Importantly, as a
process of collaboration it would bring the therapist and the client
together as a team. In other words, metaphor work would be relational
and contextual, as part of the therapy process and not
a technique to be done on
Summary and Concluding Remarks
In this paper, I explored themes of power in the accounts of clients, having analysed these accounts with a Foucauldian version of discourse analysis. A Foucauldian approach necessitates exploration of this theme as it holds that institutions like therapy maintain and reproduce power relations. Clients' constructions included metaphor as a silencing device in the hands of the therapist, whereby metaphor is used strategically, with varying degrees of success, in order to resist a particular worldview of the client and to impose another, that is favoured by the therapist. The high-jacked metaphor' is another construction that works in this way, silencing the client. As such it has the effect of alienating the client, since the meaning is not arrived at collaboratively; rather, the therapist takes the ball and runs with it (to use a metaphor), not checking out meanings with the client. This results in taking the meaning into a different direction that than intended by the client. This kind of use of client metaphors by the therapist can also take the form of another oppressive strategy that of asking probing questions and legitimises the position to be able to do that. As the therapist does not receive the metaphor as a metaphor, but rather takes it literally, the meaning again not negotiated but taken for granted by the therapist, the therapist assumes the client is talking about something literal and responds with probing psychologising questions. This has the effect of positioning the client as mentally disturbed' and has then to do a lot of discursive work to take him/herself out of that position. The high-jacked metaphor' as an unequal position of power also legitimises the asking of probing questions by the therapist and compromises the position of the client. The environment as a sign of power hierarchies and positions is another example of metaphor at work, this time not as a conscious technique used by the therapist, but as a meta-communication and an institutional method of control. In this case, the physical arrangement of the therapy space spells out the respective positions to be occupied by therapist and client. The therapist, having a bookcase behind her is expected to function as the ideal knower (a position inherently powerful) and the client as passive and emotionally discharging, as signified by the empty bin and box of tissues next to her seat. Other such signifying, metaphorical practices for power positions are objects in the therapy environment that spell out distance in the social positions occupied by the client and by the therapist, as in displayed photographs of affluence. Metaphor as meta-communication can also take the form of actions and behaviours, such as the therapist following certain scripted behaviour dictated by their profession. This can take the form of behaviours such as leading the way to the therapy room and having the client follow, taking notes and adopting a withholding stance in the process of therapy with little or no personal disclosure, feedback to the client and responses. This stance is usually consistent with the psychoanalytic paradigm, where the therapist aims to be a 'blank slate'. Metaphor, in addition to being strategically used, was also withheld by the client, as a form of reclaiming some power by eschewing a compromised positioning as 'mentally ill'. For example, I showed how some clients withheld metaphors that were anticipated to sound bizarre to the therapist, particularly in the context of the client's history of use of mental health services and the therapist's knowledge of this. A small number of instances that referred to metaphor's use as empowering were also addressed particularly as instruction/guidance and inspiration. In exploring issues of power, I also looked at client positionings that resisted or challenged unequal positions. The client as assertive was discussed in the context of education and maturity (i.e. age). Access to power was therefore deemed as a function of confidence (i.e. the ability to stand up for oneself) and knowledge (i.e. access to other more empowering discourses). Resisting compromised power positions in the absence of these factors was constructed as near impossible. Instead, the clients find themselves in a passive or even helpless position. 'Acts of resistance' constitute a number of strategies employed by the clients in order to challenge the therapist's enforcement of particular unwanted positionings. These acts of resistance' included making personal comments, asking personal questions, repetition, giving feedback, rejection, going defensive and correcting. I argue that it is not metaphor as a technique that is problematic in itself, rather it is a sole instrumentalist view of metaphor that is problematic. For example, the use of metaphor as a technique to get the client to follow instructions, such as in the case of overcoming the client's resistance (Spellman, 1994), rather as a process of meaning construction within the relational context of the therapy process. 
Anderson, H. (1997). Conversation, language and possibilities: A postmodern approach to therapy. NY: Basic Books.
Anderson, H. and Goolishian, H. (1992). The client is the expert: a not-knowing approach to therapy. In McNamee, S. and Gergen, K. (Eds.), Therapy as social construction. London: Sage.
Bhavnani, K. (1990). What's power got to do with it? Empowerment and social research. In I. Parker and Shotter, J. (1990). Deconstructing social psychology. London: Routledge.
Carswell, M. A. and Macgraw, K. (2001). Embodiment as a metaphor in therapy, Organdi Quarterly, No. 2, February, 1-12. Retrieved April, 2, 2003 from http://www.geocities.com/organdi_revue/
Cheshire, A., Lewis, D. and The Anti-Harassment Team (2004). Young people and adults in a team against harassment: Bringing forth student knowledge and skill. In D. Par‚ and G. Larner (2004, Eds.). Collaborative practice in psychology and therapy. NY: The Haworth Clinical Practice Press.
Cox, M. and A. Theilgaard (1988). Mutative metaphors in psychotherapy : The Aeolian mode. London : Jessica Kingsley.
Drew, P. and Heritage, J. (1992). Talk at work interaction in institutional settings. Cambridge: Cambridge University Press.
Edwards, D. (1997). Discourse and cognition. London: Sage.
Edwards, D. and Potter, J. (1992). Discursive psychology. London: Sage.
Gersie, A. and King, N. (1990). Storymaking in education and therapy. London: Jessica Kingsley Publishers.
Harper, D. (2002a). The tyranny of expert language, Open Mind, Vol. 113, 8-9.
Harper, D. (2002b). Moving beyond the tyranny of experts, Open Mind, Vol. 115, 20-21.
Hoffman, L. (1990). Constructing realities: an art of lenses, Family Process, 29 (1), March, 1-12.
Jaworski, A. and Coupland, N. (1999). The discourse reader. London: Routledge.
Killikelly, T. (1995). Whose dream is it anyway?, Self and Society, May, 23(2).
Kopp, R.R. (1995). Metaphor therapy-Using client-generated metaphors in psychotherapy. New York: Brunner/Mazel.
Mastache, C. (2003). Clients'
questions in therapy talk,
unpublished paper presented at the 1st International
Conference of Critical Psychology, 27-31 August, 2003, Bath.
Mastache, C. (2000). La co-construcción del proceso terapéutico: la perspectiva del paciente y la perspectiva` del terapeuta. Facultad de Psicología, UNAM, México. Unpublished MSc. thesis in Mastache, C. (2003).
Parker, I. (1992). Discourse
dynamics: Critical analysis
for social and individual psychology. London: Routledge.
Potter, J. (1996). Representing reality: Discourse, rhetoric and social construction. London: Sage.
Potter, J. and Wetherell, M. (1987). Discourse and social psychology: beyond attitudes and behaviour. London: Sage.
Silverman, D. (1997). Discourses
of counselling: HIV
counselling as social interaction. London: Sage.
Simpson, A. (1997). "It's a Game! The construction of gendered subjectivity'. In R. Wodak (Ed.), Gender and discourse. London: Sage
Spellman, D. (1994). Beyond instrumentalism: the use of metaphor in family therapy. Human Systems: The Journal of Systemic Consultation and Management, 5, 83-95.
Webster, R. (1995). Why Freud was wrong: Sin, science and psychoanalysis. London: Fontana.
White, M. and Epston, D. (1990). Narrative means to therapeutic ends. New York: Norton.
Willig, C. (2004). Discourse analysis and health psychology. In M. Murray (Ed.), Critical Health Psychology. NY: Palgrave Macmillan.
Willig, C. (2000). A
discourse-dynamic approach to the
study of subjectivity in health psychology, Theory and
Psychology, Vol. 10(4):
 A version of this paper was presented at the 1st International Conference of Critical Psychology at Bath, UK, August 2003
 Read, 'power issues'
 Names and numbers in parentheses represent a
participant's pseudonym and extract line
 FG denotes a focus group participant